View Diagrams  Download Paper  Download Images 


Whilst researching many other Portolan Charts I came across the works of Marcel Destombes and carried out an investigation of his oeuvre. Finally I came across his Document Repository and a complete, but by its descriptions, an in-substantial listing of the 26 boxes of records he had deposited. It is basically his work from 1935 to 1980 and is a massive archive. However, the compiler issued this warning; “Some notes are ambiguous (do they concern a consulted map or its description in the catalogue?) Others will no doubt remain unusable.” This listing contains 86 pages of headings and notes.

Also, in researching the texts of “Imago Mundi”, many years before, I had come across a text by M Destombes entitled, “Fragments of two medieval world maps at the Topkapi Saray Museum , dated 1955, ( It is listed in his records). I had been in touch with the Museum regarding another chart ( that took over 4 months to obtain), but when requesting a copy of these fragments, I had no response and thought the information lost to me.

However the 1955 paper became available on JStor c2008, and it contained two very grainy black and white photos of the fragments. ( I must admit to being lazy and not travelling to Paris with all the expense to obtain a copy). But, M Destombes had translated the texts on them, very few, and Hazine 1828 which is a fragment 50 x 123mm from a circular Catalan map of c200cms became so very important apropos the 1375 Catalan Atlas.

The map Hazine 1828 was to my mind obviously the forerunner of the Biblioteca Estense c1440 mappamundi. It features in the CDROM of 2004 (plus booklet) and had high resolution copies of the Estense Maps there-on. This chart however is Italian.

However, as only a small section of the northern perimeter was applicable to the extant Catalan Atlas it was possible that the residue on Hazine 1828 was applicable to the lost folio?

The actual work to find the lost folio was then put on hold for several years.

The pertinent diagrams from the original text are as follows;

Diagram 6; The Wind-Rose and thus the folio size 645 x 500mm setting out.

Diagram 7; Folios 1 and 2 ( sub-folios 1-2) set out in order to ascertain the “Normal Portolan” and its measurements to confirm it was copied from an extant Majorcan/Catalan chart.

Diagram 8; The hidden lines and score marks visible by magnifying glass set as “hypothetical latitudes” at 300 Miliaria centre.

Diagram 9: The overlap of folios 1 and 2 (sub-folios 2-3) and the mis-match in the south.

Diagram 10; The overlap or not of folios 3 and 4 (Sub-folios 6-7) and the obvious missing geography.

In this first text I made no attempt to quantify the missing folio and inserted a part of the Al-Idrisi world map (reconstruct) as an example. Obviously the Hazine 1828 fragment was intriguing but not in a state to quantify enough from the photo. Hence it was put to one side for a while. I should add that even now the TSM has not responded to several emails.


In text ChCATA/1, page 11, I wrote;
“This Atlas was drawn between c1373 and 1375 and is a labour of some magnitude. It has included the Far East and China with a multitude of cities and information appended. Read the historical texts written about the Atlas, it is a momentous undertaking of research.
But, this coastal profile could not have been drawn without a model being available. It was drawn, if all else is correct by Cresques Abraham who was an illustrator/illuminator and not a cartographer au fait with the intricacies of the chart design and layout. In other words, what he has drawn had to be copied from a cartographic paper as the texts available would not have given the correct or any real coastal profile to be so drawn. We know the sheets in the west, folios one and two are really no more than the 1339 A. Dulceto Portolan Chart onto which has been grafted the Far East and China, which must have been taken from or been prepared for copying by a cartographic draughtsman.”

That was prompted by a text posted by Juan Ceva through his excellent site, “The Cresques Project”. One text he translated was; “Cresques Abraham, Jew of Majorca; Master of Mappaemundi and Compasses” by Jaume Riera I Sans and within pages 13/14 he makes it clear just what Cresques Abraham was capable of. That text belies the title.

I quote items 1 -4 which fully set down the reality concerning Cresques Abraham.
1) “we have no record of his working on simple nautical chart and we do not see how we could assume this a priori.”

2) “With this in mind– and if one is willing to fight inertia of all that has been said and repeated about the so-called “Majorcan School of Cartography” represented by Cresques Abraham– one comes to the conclusion that the occupation of this Jew was painting or illuminating, with specialisation in the painting of compasses and mappae mundi.
Cresques Abraham would paint a compass rose with insoluble substances at the bottom of the compass, hence the name “bruixoler” ( compass maker). Cresques Abraham drafted and painted luxury mappaemundi. He did this with such ability and beauty that the royal documents would address him with the singular titles of “Master of Mappaemundi” and “Master of mappaemundi and compasses”.

3) “That Cresques Abraham was a painter is such a simple and obvious deduction that I think no one could deny it today. Furthermore, I do not think that any future document will be able to invalidate this deduction. We cannot doubt either that Cresques Abraham had mastered the art of representing geographical knowledge like no one else. However, I see it more difficult to attribute to him anything that may follow this. Today, we cannot pretend without much temerity that the occupation of “Master of Mappaemundi” in the XIV century implied more than the aforementioned skills- skills such as nautical sciences or astronomy for example.”

4) “In fact, it does not take much effort to realize that all the fantasies that have been said and repeated about the occupation of Cresques Abraham originate in the mirage of the modern meaning of the term “cartographer” that we associate with him, as well as in the ignorance of the medieval meaning of the word “bruixoler”. People pondering about the scientific knowledge of the Majorcan Jews without any prior research available, have assumed without any documental proof that the “bruixoler” Cresques Abraham “built” compasses, astrolabes and other nautical precision instruments. It has been even said– getting carried away with the fascination of the word “cartographer”– that Cresques Abraham was an astronomer, a clockmaker and a mathematician. The reality is that no medieval document or manuscript supports these allegations. No one to date has been able to demonstrate that Cresques Abraham advanced nautical sciences, the astronomical sciences or the science of cartography. However, we do know that because of his quality as an artist , King Pere and Prince Joan became two of his principal clients.”


Thus knowing that the Atlas folios 1 and 2 had to be as per the Portolan Charts available from 1313 to 1370, as my chart Diagram ChCATA/1/D19 fully explained, and given that Folios 3 and 4 were a concoction, a fairy tale taken from written texts of several centuries previous, I took the attitude that to deconstruct folios 3 and 4 was the only recourse to determine information and method.



Anybody studying the Catalan Atlas surely noted that after the first two folios, The Mediterranean Sea Basin, that the next sections were totally awry apropos the geography and the juxtaposition of Towns. Those normally not only miles apart but actually at totally different latitudes and even totally different provinces were juxtaposed. There are even duplications of towns.

To establish a basis for the exercise, the first task was to produce a geographical map with basic information which would prove a template for folios 3 and 4 and thus include the missing folio geographical area. Diagrams ChCATA/2/D01 and ChCATA/2/D02 are that map.



There-on are indicated the major geographical features, towns, routes, rivers etc., which form part of the Catalan Atlas storyline. They indicate the actual positions of towns and the territory conquered by Alexander The Great. Obviously the map does not include such areas north of China for Alexander, as a person not au fait with the actual geography could think correct.


Utilizing the geographical plot I have placed cut-out sections of the Atlas in their correct geographical locations, They are over-scaled for the base map in order that they can be easily identified and thus clearly indicate that the Catalan Atlas was prepared by a person or persons with little geographical knowledge, or a complete mis-interpretation of data.There-on are indicated the major geographical features, towns, routes, rivers etc., which form part of the Catalan Atlas storyline. They indicate the actual positions of towns and the territory conquered by Alexander The Great. Obviously the map does not include such areas north of China for Alexander, as a person not au fait with the actual geography could think correct.



From Diagram ChCATA/2/D03 it will be obvious that the majority of the movement of items to their geographical positions are from Folio 4 (sub folio7), and in fact it empties the matching geographical area. Basically in placing Alexander the Great on sub-folio 7, north of China, with what is primarily a Hindu Rite of Suttee above Lop Nor and Lake Ysicol, then setting Kabul and Camar on both sub-folios 6 & 7 clearly indicates an enormous mis-understanding of spatial positioning and a lack of co-ordination for items to be included.

Also, on sub-folio 7, the Mounts of Baldassia mimic the Mounts of Amol and are possible mis-named as the Mounts of Amol are part of the Elburz Mounts of N Persia and are on the southern section of the Caspian Sea. The Mounts of Baldassia are part of the Pamirs of Kyrgistan.

Perhaps the biggest error in placement is that of the Indus River, because if the Mounts of Baldassia are correctly placed and inverted, then the river Indus appears flowing south. Then, with a city of Alexander placed in India N E of Delhi with the Desert of India, the Khar or Great Indian Desert, east of Delhi, it leads to the opinion that from sub-folio 7 west to sub-folio 6 should in fact be handed. It is drawn from an upside down perspective and should be read from 7 to 6 being East to West? ( The reason for this is quantified later).
But as Diagram ChCATA/2/D04 clearly indicates there is no part of the Catalan Atlas which can be considered as filling the great void on the geographical map base.


Having seen the Destombes black and white photos of the Topkapi Saray Museum fragments of the circular maps, Hazine 1827 and Hazine 1828, I originally wrote this following section of text featuring that which I could glean from those photographs and the accompanying text. Diagram ChCATA/2/D10. Having completed that text, I was contacted by Professor Ramon J Pujades I Bataller in response to a question regarding the place names on the C41 chart, c1445, Florence BNC Port 16, Anon Majorcan (Vallseca’s ateletier) 92 x 131cms. I had obtained a copy direct from the Library even though it was freely available on-line and told that on-line was also the complete scan of the library catalogue, a hand written booklet which contained a transcription of the texts there-on. However I could see that there were errors and a confusion there-in and sought guidance, as it appeared two or three texts were missing. There was a wait before the response appeared and it was far greater than I could ever imagine. Professor Ramon J Pujades I Bataller had placed the day before his response, on-line a new book, “Late Medieval Mappamundi; From the Birth of the Hybrid to the Demise of the Portolan and Mappamundi”.

There-in each Catalan and Italian map/chart as well as written texts are transcribed and translated for all to use and Professor Pujades actually wrote to me, that I could send it to any researcher I thought would appreciate it. Hence I placed a note on my Academia Edu., page to that effect, giving the URL for access. In doing so Professor Pujades had provided me with the coloured copies of the two fragments from TSM, Hazine 1827 and 1828 and a complete transcription/translation of the texts there-on. Hence, this section is a complete re-write.


This section of a large circular chart is dated 1370/1380 and hence is quite possibly pre-Catalan Atlas of c1374/75 and thus the “Mappamundi” that the Cresques Abraham atelier used to produce the Atlas. Professor Pujades writes on page 490;
“Upper fragment of a large circular portolan mappamundi, surrounded by an outer ocean and covered in a rhumb line network. A high-luxury piece, decorated with an abundance of miniatures and other ornamental motifs, and enriched with a number of legends. Its craftmanship, colour scheme and characteristics are extraordinarily similar to the ones described for the Catalan Atlas. By the same token, it contains precisely the north-eastern sector of the cartographic design of the oecumene, namely, the northern part of the Asian continent and of eastern Europe. It thus includes the part that could not fit in the upper area of panels II and IV of the Catalan Atlas due to lack of space, with which it matches up nearly perfectly along the southern edge.. Since it seems written and painted by the same hands that wrote and decorated the Catalan Atlas, both its attribution to the Cresques Abraham Atelier and its dating to the 1370’s give little cause for doubt.”


I cannot subscribe to it not fitting panels III and IV of the Atlas as it is clear that there is a panel missing and thus caution would be advised here.

However, the attribution to the Atelier is excellent and indicates that there was a person there-in capable of producing/copying a Portolan Chart and understanding its methodology to draw it precisely and thus probably not just a Painter/Illuminator.


This fragment is attributable to Bartolomeo de Pareto atelier, Genoa, c1455. The chart, BNC.CN1 (c57) attribution “Presbite Bartolomeus de Pareto, civis Janue, acolitus sanctissimi domini nostri pape, composut hamc cartam MCCCC°LV in Janua”.

The fragment is of a rectangular framed circular mappamundi and is 50 x 85cms, representing about a quarter, the N E part, of a circular map inside a decorated frame. Hence we are looking at the third circular world map, the second being the c1440 Biblioteca Estense mappamundi.


Place all three together and it is obvious they stem from a basic original, but all are variants of it in text and diagrams there-on. That probably indicates a mappamundi earlier than the Hazine 1828 Cresques Abraham Atelier version and could have been in development from 1339 onwards, on Majorca, perhaps culminating in the C A Atelier version.

That fact does not bode well for the task of determining a putative missing folio as it pushed the dates for knowledge earlier with less available of course.


Diagrams ChCATA/2/D07 and D08 con-joined.
The A2 diagram overlaps centrally as two A3 sheets and is bounded by sub-folios 6 and 7 with thus a central section correctly sized for the missing folio. In the centre north part of Hazine 1828 circular chart has been so positioned that the cremation cartouche is in the correct position in the east.

The second inlet in the north is set centrally to the Septentrional Star and there-on I have numbered the texts and appended them for reference.

Because the “terrain” of Alexander the Great traverses this new folio to join the sub folios 6 and 7 “Alexander texts”, I have included on the geographical layout the cities on the Indus (D01/D02) which could well be spread across the folios. In the south is included a map of the Brahmaputra and Ganges rivers which a re missing from the original and I have included the overlay Mounts of Baldassia and the Indus River geographically positioned.

ChCATA/2/D07 & D08

On sub-folio 7 each item which should be on sub-folio 6 is identified with only Lop Nor, Isykol Lake and Karachoja of the Takla Makan Desert along with Caracora/Karakorum, the Mongol Capital of Kublai Khan on the correct eastern side, although set basically upside down south to north. That probably indicates, yet again, that the map being used to copy from has been reversed and thus mis-read.

The N W , sub-folio 6 has the caravan travelling from Sara to Cathay and above it is the city of Camul, noted as the “Last city of Cathay”. Hence , we have Alexander on sub-folio 7, with Chabol and Camar set above Cathay and yet Cathay is supposedly also north of Tashkent, which itself is south of both Russian Siberia and Kazakstan. Until China annexed Tibet, the Xinjiang Province was the western extremity of China with Urumchi to the north and Kashgar in the west, the chief city of Chinese Turkistan and is c7 1/20 longitude east of Tashkent, not due north.( Having visited Urumchi, Takla Makan etc., these facts are correct.)

The cities surrounding the River from the Caspain Sea to the Mounts of Amol are reasonably correctly positioned apropos each other and thus the river is either the Jaxartes or the Oxus, terminating at the Amol Mts.

As already stated the River Indus and its tributaries are missing, Alexanders territory!
The SW has the misplaced “Desert of the Indies” and a city of Alexander as well as thus Ciutat de Delhi. However, the west coast of India, although truncated, does roughly accord to the geography apropos the towns.

In the east, sub-folio 7 is basically full of mis-placed towns, which if listed out appear to be more likely set reverse, N/S, with Chabul and Camar adjacent to the Mounts de Baldassia and to their east Carachora, Karakorum, Kublai Khans capital, but very curiously ELBUT, which is Tibet?

Including the Illa Jana set against the Indian Sub-continent, sub-folio 7 is basically nonsense and clearly indicates no coherent positioning of towns etc., or copying from Al-Idrisi?


The atlas is a visual feast for the eyes and as a gift from one King to another obviously intended to have the “wow factor”, rather than be a geographical treatise. That it certainly has and given that the only section which the recipient King would have any knowledge of, “The Mediterranean Sea Basin”, that being expertly drawn, would have given credence to the other folios.

Thus folios 1 and 2 would be accepted without question and other folios (extant with the missing folio then) taken to be correct through lack of actual geographical knowledge.

Africa was explored internally for trade c1325 onwards but the West coast, even though Jaume Ferrer’s voyage in search of the “River of Gold”, c1346 is noted on the atlas, in fact the main Portuguese explorations commenced c1434CE. Hence it was an aberration and added no knowledge for the production of charts.

The Indian Ocean is explored by Arab traders, but it is not until c1414 and again in 1487 that the Europeans penetrated the area and this included India, Ceylon, The Golden Chersonesus, Taprobane, and Java. Thus c 1370 there was no geographical evidence to be used to concoct the coastlines and island positions, except Al-Idrisi..

However the Great Silk Roads from Constantinople/Istanbul to Xian were known of with the routes being used from c100BCE to 1650CE, and these were known by lengths of days travel. That of course varied considerably depending upon mode of travel. The other Silk Roads and Sea routes dated 1245-1295CE were known and thus Marco Polo 1271CE traversed Persia, then skirting the Hindu Kush to Kashgar, Khotan and south to the Takla Makan and finally Shang-Tu and Khanbalik (Beijing).

In 1245 John of Plano Carpini, a Catholic Archbishop of Serbia (1247), left to travel to Karakoram to meet the Great Khan. However his manuscript is short on geographical description and he actually failed in his quest. Then in 1253 William of Rubruk makes a similar journey arriving in Karakorum on Palm Sunday 1254. John of Plano Carpini does however mention the land of Comania and lists the various peoples as he states it is a large and long country through which he travelled to the land of the Kangittae, then the Bisermini, with its Lord “Golden Alti”. He states, ”on the south side it hath Ierusalem and Baldach and all the whole country of the Saracens”. He also mentions the rivers that were crossed which indicate his route. The Neper, Tanai, Rha and Rhynus being the Borysthenes, Don, Volga and Iaec. Basically these two explorers used the route from Sarai, north of the Caspian and Aral Seas, then along the Jaxartes and north of the Tien Shan mountains towards Karakorum.

Thus textual knowledge there was aplenty but no real spatial knowledge, the juxtapositions, NSEW. And real distances that would have ensured a donor circular map used to draw the Catalan Atlas had sufficient accuracy to obviate the problems that are contained on the folios.

The text of Marco Polo first published in 1301, “Description of the World” has a fuller set of data in later editions. The “POLO’s” departed 1271 for Tartaria and then in 1292 returned to Italy. The text is full of place Names, now and again with day travel distances but it also lacks spatial knowledge for producing a coherent map.

We cannot therefore, through hind-sight criticize the Atlas for its intent, merely the lack of co-ordination in its production.

In my original text ChCATA/1, I utilized a section of the Al-Idrisi world chart as produced by Konrad Miller from the original 70 single sheets. The reason for its choice being the form of the mountains and coastlines which contain a great similarity to the Estense mappamundi and hence also Hazine 1828.



I also noted that the Al-Idrisi chart followed the Arabic methodology of having South to the top of the sheet. Hence those maps are basically SNWE, when the European maps are NSEW and that to me was a marker for the mis-placement of the cities etc., on folios 3 and 4. The text of Al-Idrisi ( I use a French translation) and studying the first three climates and the 30 co-joined sheets to the Persian Gulf and the Eastern Sea, there are major cities listed and positioned on the sheets. The date of the Al-Idrisi text , even though it is in Arabic, given the population of Majorca would have been ideal for the “Cresques Clan” to utilize.

Finally , this text has A LOT OF NEGATIVITY AGAINST THE Eastern portion of the Atlas, discussing even what the missing folio may actually have shown, because folios 3 and 4 are such a mixed up presentation that it is hard to grasp how the information obtained, that is of course if Al-Idrisi was used, was so mis-interpreted and therefore lead to placing of the cities so haphazardly on those folios. It is hard to reconcile the Al-Idrisi folios which are obviously upside down for Europe not being upside down for Asia.

That must not detract from the beauty of the Catalan Atlas, no doubt why the “Cresques Clan “ were chosen, being superb painters and illustrators, but “masters of mappaemundi” they are not.

But, the last comment must be, if Hazine 1828 is the precursor to the Catalan Atlas it is likely that Hazine 1827 was either copied by an Italian in Majorca or from a copy sent to Italy after the Catalan Atlas was completed in 1375. That raises the question of their journey from Italy to Turkey and when they were so badly mutilated? Answers perhaps another day!

But, it also raises the question, already posited, “is Hazine 1828 the original map”?